8 Psychological Article Summaries on Father/Infant Attachment Belsky, Jay, Cambell, Susan, & Cohn, Jeffrey. (1996). Instability of infant-parent attachment security. Developmental Psychology, 32, 921-924. This relatively simple study examined the stability of infant-parent attachment over a period of approximately six months. The authors stated that previous researchers found stability rates as high 96%, but inferred that these studies might have been flawed because of small sample sizes and differences in the way the Strange Situation was scored. This is the first study to investigate the stability of infant-parent attachment in nearly a decade and though stability was thought to still occur at a significant rate, the authors suggested that the rate of stability would probably occur at a lower rate than has previously been suggested. Two samples of participants were used in this investigation. The Pennsylvania State University (PSU) sample consisted of 125 middle class, caucasian families who reared first born sons. The participants in this sample were originally recruited as part of a study investigating the effects of family stress on infants. Though this sample consisted of only male infants, this was not thought to be an issue because previous research has failed to find sex differences in regards to attachment stability. The University of Pittsburgh (Pitt) sample consisted of 90 infant-mother dyads who were originally part of a study on post-partum depression. The mothers in this study were middle class and first time parents. Nearly one-half of the mothers were diagnosed as being clinically depressed after the delivery of their infants, though 80 percent of these depressed participants had remitted by the time their infants were 6 months of age. To determine infant-parent attachment, all of the infants were observed in the Ainsworth and Witting Strange Situation with their mothers at the ages of 12 and 18 months. In the case of the PSU sample, the infants were also observed in the Strange Situation with their fathers at the ages of 13 and 20 months. Infants were classified as either insecure-avoidant, secure, or insecure-resistant in regards to their attachment with either their mothers or fathers. Attachment stability was determined by calculating coefficient lambda. This assesses the degree to which the infant-parent attachment style at a later date can be predicted by the knowledge of the infant-parent attachment style at an earlier date. Coefficient lambda can range from 0.0 (no predictability) to 1.0 (perfect predictability). As stated previously, earlier studies have found coefficient lambdas as high as .96. The coefficient lambda for the Pitt sample was .46 for infant-mother attachment. For the PSU sample they were .52 for infant-mother attachment and .46 for infant-father attachment. None of these results were significant. Surprisingly, the results of this study seem to suggest that neither infant-mother nor infant-father attachment styles remain stable during infancy. This is quite different from what previous researchers have concluded. Concerned by the prospect that the larger proportion of depressed mothers in the Pitt sample might have influenced the results, Belsky et al. analyzed that sample’s stability using only the non-depressed participants which produced no significant difference from the first analysis. The authors suggested that either past researchers have greatly overestimated stability in attachment or some environmental change has occurred in our society, during the past 10 years, that is effecting the stability of infant attachment. They mentioned that a majority of the employed mothers in both samples returned to work before the end of the infants first year of life and that this might have effected the results due to the change in infant-mother relations. Also mentioned was that fathers appear to be taking an increased role as parent and this might also effect both infant-mother and infant-father attachment stability. In conclusion, Beslky stated that infant-parent attachment should be viewed in terms of instability rather than stability. Overall, I thought this was a useful study because it appears to contradict past views of infant-parent attachment stability. Perhaps it will spawn more research into this seemingly neglected topic. Though this study has found that infant-parent attachment appears to be rather unstable, I found Belsky’s explanations of why this occurs to be lacking. As previously stated, he suggested that stability might be effected by the mothers quick return to the workforce. I find it perplexing that he did not do some sort of post-hoc analysis on this dichotomy to determine if the mothers who did not return to work as quickly differed, in terms of attachment stability, from the mothers who did return to work quickly. He had the means to pursue this, and unless I am forgetting an important aspect of statistical analyses that would prevent this from being valid it seems as though this should have been pursued. He also could have conducted some sort of follow up interview with the pa! rents to determine if their was something different between these samples and the samples of previous investigations. Besides these limitations, the overall validity of the study appears to be acceptable. Hopefully more studies will investigate this phenomenon further. Belsky, Jay. (1996). Parent, infant, and social-contextual antecedents of father-son attachment security. Developmental Psychology, 32, 905-913. This study examined various personality traits and social-familial characteristics that might influence infant-father attachment. 126 father-infant dyads were used in this study. All of the infants were first born, caucasian males. The average age of the fathers was 31 years and they had, on average, 15 years of education. Their average yearly income was $40,000. Fathers completed several questionnaires to assess the antecedents of infant-father attachment when the infants were 10 months of age. To assess the personalities of the fathers the NEO Personality Inventory was administered. This assesses personality in terms of extraversion, neuroticism, and agreeableness. Other tests administered to the fathers assessed the quality of their marriage, the amount of social support they believe they receive, the degree to which their work interferes with their family and vice-versa, and their perceptions of their sons’ personalities. Infant-father attachment was assessed, three months later, by a videotaping of the fathers and infants in the Strange Situation. The resulting data was analyzed using a MANOVA that used the scores from the various tests the fathers were given to predict the infant-father attachment style. The results suggest that fathers of securely attached infants, when compared to fathers of insecurely attached infants, tend to be more extraverted and agreeable, tend to have more positive marriages, and tend to experience more harmony and support across both work and family life. The results also suggest that fathers of insecure-avoidant sons tend to perceive their sons’ temperaments as being more positive than do fathers’ of insecure-resistant sons. Belsky states that because the roles of mothers and fathers are continually changing, it is important to investigate how fathers influence their offspring’s personalities and behaviors. More fathers than ever before are the primary caretakers of their infants and thus it definitely is important that we understand how they can positively and negatively effect their children. The real limitation of this study was that it only studied infant males, instead of both males and females. According to Belsky, “the exclusive focus on sons in this report was based in part on evidence that fathers are more involved with sons than daughters and that by early in the second year sons prefer their fathers as playmates.” Fortunately previous attachment researchers have not shared this point of view because there would not be any information on mother-son attachment if they had. It is my opinion that another reason only sons were used in this study is because the most convenient sample ! happened to only contain sons. There is nothing particularly wrong with this approach, due to the extreme difficulty in acquiring a large sample of father-infant dyads, but it definitely limits the generalizability of the results. Cox, Martha, Henderson, V., Margand, Nancy, & Owen, Margaret. (1992). Prediction of infant-father and infant-mother attachment. Developmental Psychology, 28, 474-483. This study investigated several possible correlates of infant-parent attachment. Specifically, the researchers investigated whether the behaviors and attitudes of mothers and fathers, towards their infants at an early age, influence their infants’ attachment to them later in life. This investigation was part of a longitudinal study that was to continue for approximately six years. The participants consisted of 32 caucasian husband-wife-infant triads, of which the parents’ mean education length was approximately 15.6 years, age was 28 years, and annual family income was $37,000. Infant-parent attachment was assessed at three months and 12 months of age. In each of these intervals the infants were observed in Ainsworth’s Strange Situation, first with one parent, then seven to 10 weeks later with the other parent. Infant-parent attachment was measured on a scale ranging from 1 (least secure) to 4 (most secure). Each parent was also interviewed separately for approximately one hour immediately following the attachment assessment. These interviews assessed several dimensions of the infant-parent relationship including the parents’ attitudes towards their infants, the amount of time they spend with their infants, and the degree to which the interactions with their infants are positive. In combination with these interviews, each parent was also separately observed with his/her infant. During these observations, observers recorded various behaviors that occurred between the parent and infant, and also rated the infant-parent interaction in terms of i! ts quality. The researchers were interested in knowing whether certain behaviors and attitudes parents have towards their children at three months of age has an effect on those children at 12 months of age. To accomplish this, the parents’ attitudes and behaviors, during their children’s third month of life, were used to predict the degree to which their children would be securely attached at 12 months of age. The results suggest that mothers who positively interact with, show more physical affection towards, and spend more time with their infants will be more likely to have securely attached infants than those mothers who exhibit opposite characteristics. The results continue to suggest that fathers who positively interact with their infants and have a generally positive attitude towards their infants and their parental role will be more likely to have securely attached infants. The real importance of this study is that it addresses the role that fathers play in the development of securely attached infants. As the results indicate, fathers do appear to play somewhat of a role in this area, but in different ways than do mothers. For example, mothers’ physical affection towards their children seem to be very important to the development of securely attached infants whereas fathers’ physical affection towards their children only play a minor role. Even more surprising, infants who’s mothers spend more time with them seem more likely to become securely attached, whereas infants who’s fathers spend more time with them are actually less likely to become securely attached when compared to infants who’s fathers spend less time with them. The authors suggested this can be explained by inferring that infants who’s fathers spent more time with them also had mothers who spent less time with them. This compensation for less time spent with their mothers co! uld be detrimental to the development of a secure attachment according to the authors. I believe this speaks to the importance of the infant-mother relationship and to the possibility that infant-father relationships are not as important to the development of the infant. Steele, Howard, Steele, Miriam, & Fonagy, Peter. (1996). Associations among attachment classifications of mothers, fathers, and their infants. Child Development, 67, 541-555. These researchers investigated the relationships between infant attachment, father attachment, and mother attachment. Significant relationships between these measures would suggest that parents’ adult attachment styles play a role in determining the attachment of their infant. The researchers also examined possible interaction effects of parents’ attachment styles on the attachment style of the infant. The participants of this study included 90 caucasian, middle class infant-mother-father triads, of which the parents’ average age was 31. 47 of the infants were boys and 43 were girls. Each parent was administered the Adult Attachment Interview in order to assess the parents’ attachment style. All of the interviews were independently rated by four trained judges. Kendall’s kappa calculations were performed on the judges’ ratings and they were found to be highly reliable. Based on the interviews, the parents were classified as either insecure-entangled (E), insecure-dismissing (D), secure-autonomous (F), or unresolved (U). Infants were also classified as to their attachment styles by the administration of the Ainsworth Strange Situation. They were classified as either avoidant, resistant, secure, or disorganized. The data was first analyzed to determine if the infants’ attachments to their mothers was similar to the attachments they had to their fathers. Kendall’s kappa calculations were significant for this analysis, meaning the infants’ attachments to their mothers and fathers tended to be similar. In order to investigate whether infants’ attachment styles were similar to that of their parents the researchers fist had to decide which infant attachment styles were most similar to which adult attachment styles. What they decided was that securely attached adults were most similar to securely attached infants, insecure-dismissing adults were most similar to insecure-avoidant infants, insecure-unresolved adults were most similar to insecure-disorganized infants, and insecure-entangled adults were most similar to insecure-resistant infants. The strength of the relationship between these four pairs of measurements was then calculated using Kendall’s kappa. The results suggest that,! in fact, mothers and fathers both tend to have similar attachment styles as their infants. Finally, chi square analyses were performed to investigate possible interaction effects between mother and father attachment styles. The researchers suggest that these results indicate that the mothers’ attachment style is more important than the fathers’ attachment style in determining the infants’ attachment. This study reveals that mothers and fathers appear to influence their children’s attachment in a way that makes it similar to their own adult attachment. Of course it is also possible that infant attachment is actually influencing parent adult attachment. This research design renders it impossible to establish the direction of the effect. Further research needs to be done in order to investigate the direction and reasons behind this phenomenon. Fox, Nathan, Kimmerly, Nancy, & Schafer, William. (1991). Attachment to mother / attachment to father: a meta-analysis. Child Development, 62, 210-225. The use of a meta-analysis permits researchers to examine relatively large samples by combining the results of several different experiments. Often the combined results of several experiments provide a clearer understanding of the issue in comparison to just interpreting a single study’s results. This technique is of particular importance when various experiments provide conflicting and sometimes opposite results. By pooling the results of several studies, researchers are able to make a more informed and complete decision as to what past studies appear to suggest. This study employs the use of a meta-analysis to examine the combined results of 11 studies examining similarities and differences between infant-mother and infant-father attachment. The authors of this study found that 11 experiments have been conducted on infant-mother and infant-father attachment. The authors of these 11 studies were contacted and each of them gave the authors of this current study the ra! w data and other information about the samples used in their studies. In all of these studies, each parent was observed separately, with their infant, in the Ainsworth Strange Situation in order to assess infant-parent attachment. In nine of the 11 studies the infants were first observed when they were 12 months of age. In the other two they were first observed when they were 11 months of age. In eight of these studies the order in which parents were observed with their infants was counterbalanced. In the remaining three studies all of the mothers were first observed with their infants, followed by the fathers. The time intervals between observations ranged from one week to six months, with a mean of 11.2 weeks. In all of the studies the coders of either the infant-mother attachment or infant-father attachment were blind to the infants’ attachment to the remaining parent. Four of the studies used only firstborn infants, though the remaining studies found no influence of birth order on infant-parent attachment. Nine of the studies used! a similar number of male and female infant participants. Of the remaining two, one used significantly more infant males than females and the other used significantly more infant females than males. The number of infant participants used, among all 11 studies, ranged from 32 to 132, with an average of approximately 60. Six analyses were completed in this study that assessed the agreement between infant-mother attachment and infant-father attachment. In all of the analyses Kendall’s kappa was calculated to determine the level of agreement. The first analysis assessed the level of agreement between secure and insecure infant attachment between mothers and fathers. A kappa level of .309 was obtained for this measure suggesting that infants who were securely attached to one parent tended to be securely attached to the other parent, with the same suggestion for insecurely attached infants. The second analysis assessed the level of agreement between insecure-avoidant and insecure-resistant infant attachment between mothers and fathers. A kappa level of .734 was obtained for this measure, suggesting that infants who were insecure-avoidantly attached to one parent tended to be insecure-avoidantly attached to the other parent, with the same suggestion for insecure-resistantly attached infants! . The third and fourth analyses produced similar results when infant-mother and infant-father attachment was compared for infants in various sub-categories of the three categories of attachment. This suggests that infants who are classified as having a certain sub-category of attachment towards one parent tends to have the same sub-category of attachment towards the other parent. Unfortunately the authors of this study did not state what each of the sub-categories of attachment are. I am therefore unable to explain in more detail what the third and fourth analyses should suggest to the reader. The final two analyses explored the hypothesis that the results of the first four analyses were not suggesting that infants who are attached in a certain way to one parent are also attached similarly to the other parent, but instead were suggesting that infants who are more likely to cry with one parent are also more likely to cry with the other parent. The analyses supported the! hypothesis and suggest that many of the 11 studies used in this meta-analysis were flawed because the researchers might not have been properly classifying the infants according to their attachment, but instead according to their temperament. This study was very useful because it seems to provide fairly substantiated evidence that infant-parent attachment is dependent upon both parents. The limitation of this study is that it cannot provide an explanation for how or why infants are similarly attached to both parents. It is possible that one of the parents is the person who the infant truly develops an attachment style towards and when the other parent is observed with the infant, the infant is simply projecting that learned attachment towards her/him. It seems likely that most infants would develop an attachment style towards her/his mother because she is probably the person who does most of the caretaking. So in this example the attachment displayed towards the father would simply be a duplicate behavior with far less importance and substance. On a topic that is somewhat unrelated to my original topic of infant-father attachment, the authors of this article did make an important contribution when they sugg! ested that the 11 studies used in this meta-analysis might be classifying the infants according to their temperament instead of their attachment. It seems difficult to separate infant temperament from infant attachment because they both appear to influence one another. It is likely that infants who are quicker to cry would be classified a certain way, but it is also likely that infants who have certain attachment styles are quicker to cry. This might represent a flaw in the methods that are used to establish infant attachment styles. Perhaps methods that do not rely so heavily on emotional displays of anger and sadness would lessen the extraneous effects of infant temperament. Ferketich, Sandra, & Mercer, Ramona. (1995). Paternal-infant attachment of experienced and inexperienced fathers during infancy. Nursing Research, 44, 31-37. These researchers were interested in studying various factors that influence father-infant attachment. Specifically, they wanted to know whether or not inexperienced fathers were influenced by different factors than were experienced fathers. In order to investigate this the researchers conducted a longitudinal study of two samples of fathers; first-time fathers and experienced fathers. Both samples were assessed in terms of their attachment to their infants and various others beliefs they held towards their infants. These assessments occurred four times over the course of the infants first eight months of life. The participants consisted of 79 experienced fathers and 93 inexperienced fathers. Experienced fathers were significantly older than the inexperienced fathers by an average of approximately four years, but the two samples did not differ on any other demographic dimensions. The majority of these two samples were caucasian, had a professional socioeconomic status, and were married to the mothers of their child(ren). Of the experienced fathers 65% had one previous child, 23% had two previous children, and the remaining 12% had between three and eight previous children. Infant-father attachment was assessed using the How I Feel About My Baby Now Scale. This is a 10 question survey with item responses ranging from one to four, thus yielding a total score between 10 and 40. Paternal competence was assessed using the 17-item Parenting Sense and Competence Scale. Self-esteem was measured using the 10-item Rosenberg Self-Esteem Scale. The father-mother relationship was assessed using the Locke and Wallace Marital Adjustment Test. Family functioning was measured using the 21-item Feetham Family Functioning Instrument. Depression was assessed using the 20-item Center for Epedimiologic Studies Depression Scale. Anxiety was measured using the State Anxieties Scales. Social support was measured using the 40-item Inventory of Socially Supportive Behaviors. Stress was measured using the Norbeck Life Experiences Survey. Fetal attachment was measured using the 23-item Cranley Fetal Attachment Scale. Several other single-item measures were als! o taken on various aspects, including the father’s perception of the infant’s health and delivery experience. As stated previously, the fathers were assessed, in terms of the above measures, four times during this study. Measures were taken immediately after the infants were delivered, one month after delivery, four months after delivery, and eight months after delivery. A multiple regression analysis was conducted, using the various measures as predictors of infant-father attachment. The results suggested that, for experienced fathers, fetal attachment was positively correlated with attachment during the first four months. Either depression or anxiety were negatively correlated with attachment at each of the four testing dates. A positive labor and delivery experience was positively correlated with attachment at one months and eight months. Finally, at one month, a positive relationship with the mother was negatively correlated with attachment, and the amount of contact with the infant was positively correlated with attachment. For inexperienced fathers, fetal attachment was positively correlated with attachment during the first two months. Depression was also negatively correlated with attachment at two and four months. Furthermore, inexperienced fathers who felt they were competent as fathers also tended to have higher levels of a! ttachment immediately after birth and at eight months. Finally, there were no differences between experienced and inexperienced fathers in terms of their overall level of attachment at any of the four testing dates. This study provides more insight into the workings of father-infant attachment. What the authors state is that fathers develop unique loving relationships with each child they have. This explains the similarity of attachment when experienced fathers are compared to inexperienced fathers. Furthermore, they state that depression tends to have a negative effect on attachment for both groups of fathers. This is a serious problem according to the authors and should be addressed by hospital counselors because depression is not an uncommon experience for fathers and mothers following the delivery of their child. Competency was a significant predictor of attachment for inexperienced fathers which might reflect a lack of confidence, in their parenting ability, by inexperienced fathers that is not experienced by experienced fathers. This is not so unusual because in areas other than fatherhood, experience breeds confidence which sometimes leads to more involvement. This was a useful article because it shed some light on factors that influence paternal attachment. This seems to be an overlooked, though important topic by many attachment researchers. Past research appears to conflict with one another in terms of the influence and importance of paternal attachment with regards to the development of the infant, but without conclusive evidence of its a complete lack of influence more research such as the present study needs to be completed on this topic. Volling, Brenda, & Belsky, Jay. (1992). The contribution of mother-child and father-child relationships to the quality of sibling interaction: a longitudinal study. Child Development, 63, 1209-1222. This study was conducted in order to investigate how various aspects of the parent-child relationship influence sibling interaction. This is related to my topic because one of the aspects of the parent-child relationship that was studied is attachment and because it dealt with the attachment of fathers as well as that of mothers. This article documents results that were part of a six-year longitudinal study. The participants included 30 maritally intact, caucasian families. Each of these families had one sibling pair that consisted of brother-brother, sister-sister, or brother-sister pairs. The mean ages of the older and younger siblings were 72 months and 39 months respectively. The mean age difference between the older and younger sibling was 32 months. The fathers and mothers of the families were, on average, 34.6 and 32.6 years of age respectively. The mean family income was $33,454. All of these families were first recruited for this study during the pregnancy of their first born child. There were too many different measurements taken during this study to be presented in this summary. Instead, only the measurements that are related to my topic will be discussed. Infant-parent attachment was assessed when the infants were one year old and three years old. For the one-year measurement the Ainsworth Strange Situation was used, using both mother-infant and father-infant assessments separated by one month intervals. For the three-year measurement the children were observed separately with their mothers and fathers in the laboratory. They were observed in seven different activities and five raters rated them in terms of various qualities related to attachment, including parent-infant affect and parent responsiveness to their children. Sibling interaction was also observed and rated in terms of conflict, prosocial behavior, and shared affect. It appears that sibling interaction observations occurred when the first born child was six years of age. These ob! servations occurred twice and were two to three weeks apart. The results of this study suggest that sibling conflict was greater for children if they had an insecure attachment to their mothers at one year of age and if their mothers were intrusive and controlling at the children’s third year of age. Prosocial sibling interaction was increased when the fathers were supportive when the children were three years of age. It was decreased when the fathers were either intrusive or uninvolved when the children were three years of age. An interesting finding of this study was that infant-father attachment did not seem to effect sibling conflict, but instead effected prosocial sibling conflict. According to the authors, these findings are consistent with previous researchers’ findings. One problem with this study was that different measures of attachment were used at the two ages when it was assessed. The reason for differing measurements is assumed to be because the children being measured had aged apparently beyond the scope of the Ainsworth Strange Situation, and thus forced the researchers to use a more age appropriate measurement device. Albeit, there still may be conflicts between the two measurements that could have effected the results. I did not see any mention of this anywhere in this article, therefore I am not convinced that these two measures can be used interchangeably. I have noticed that many of the articles that Jay Belsky authored or co-authored seem to be somewhat difficult to understand and comprehend. Perhaps it is because the issues that he deals with are very complicated, but I must say that some of the confusion with his writings could have been resolved with a more detailed and complete explanation of what he did and how he came to all of his conclusions. This being said, the article still provided some valuable insight into the effects of both mothers’ and fathers’ attachment to their infants. Establishment and Effects of Infant-Father Attachment J.D. Foster Proseminar II April 14, 1998